mirror of
https://github.com/scylladb/scylladb.git
synced 2026-05-28 10:41:12 +00:00
storage_proxy: stop waiting for hints replay when node goes down
Now, when repair coordinator is waiting for hints to be replayed on some remote node and that node goes down, it stops waiting for it.
This commit is contained in:
@@ -5342,6 +5342,12 @@ future<> storage_proxy::wait_for_hints_to_be_replayed(utils::UUID operation_id,
|
||||
// Sleep 1 second between checks
|
||||
bool reached_sync_point = false;
|
||||
while (!reached_sync_point && lowres_clock::now() < deadline) {
|
||||
// Check if the destination is alive
|
||||
if (!gms::get_local_gossiper().is_alive(addr)) {
|
||||
slogger.debug("Node {} is no longer alive, won't wait anymore for its hint sync point", addr);
|
||||
break;
|
||||
}
|
||||
|
||||
slogger.debug("Waiting for all hints from endpoint {} to be sent out; time spent so far: {}s, targets: {}",
|
||||
addr, std::chrono::duration_cast<std::chrono::seconds>(lowres_clock::now() - start_time).count(), target_endpoints);
|
||||
try {
|
||||
|
||||
Reference in New Issue
Block a user